TPSTokens Per Second

Gemini 2.0 Flash

Google DeepMind🇺🇸 United States
active

Google's fastest production model with 1M context and native multimodal I/O.

Compare with other models →
Context window1000K tokens
Input / 1M tokens$0.1
Output / 1M tokens$0.4

Version History

gemini-2.0-flash-001major

GA release with native image generation, live audio, and 1M token context.

2.0-flash-001major

Gemini 2.0 Flash introduces significantly faster time-to-first-token compared to Gemini 1.5 Flash while maintaining quality parity with larger models. Enhancements target multimodal understanding, coding, complex instructions, and function calling.

Benchmark Scores

Full leaderboard →
1360.0 elo
Arena Elo
86.5%
Hallucination (Knowledge)
228.0 tokens_per_sec
Speed (tok/s)